In te n sity (Cou n ts)

INCORPORATION OF CARBON NANOTUBES IN In₂O₃ THIN FILMS FOR GREATER DSSC PERFORMANCE

Savisha Mahalingam¹, H Abdullah¹, Sahabudin Shaari², and Andanastuti Muchtar³

1Department of Electrical, Electronic & System, Faculty of Engineering and Built Environment,

Universiti Kebangsaan Malaysia, 43600 Bangi, Selangor, Malaysia

2Institute of Microengineering and Nanoelectronics (IMEN), Universiti Kebangsaan Malaysia, 43600 Bangi, Selangor, Malaysia

3Department of Mechanical and Materials Engineering, Faculty of Engineering and Built Environment,

Universiti Kebangsaan Malaysia, 43600 Bangi, Selangor, Malaysia.

Corresponding author: huda@eng.ukm.my

ABSTRACT

In₂O₃ is a wide band gap material which has a potential application in dye-sensitized solar cells (DSSCs). The objective of this research is to enhance the photovoltaic conversion efficiency in In₂O₃ thin films by incorporating MWCNTs and SWCNTs in the nanocomposite. In₂O₃ incorporating CNTs annealed at 450 °C was prepared by using spin-coating method for DSSCs. The structural and morphological characteristics of In₂O₃ thin films were studied via XRD patterns and atomic force microscopy (AFM).

The In₂O₃-SWCNTs based DSSCs exhibited better photovoltaic performance than In₂O₃-based DSSCs with a conversion efficiency of 1.11 %. The electrochemical impedance spectroscopy (EIS) unit investigated the charge transfer process inside the DSSCs. A longer electron lifetime with a high recombination rate was found in In₂O₃- SWCNTs. These two factors speeds up the electron transport in the cell and enhanced the PCE of the DSSCs.

Keywords: In2O3-MWCNTs; In2O3-SWCNTs; morphology; electrical; DSSCs;

INTRODUCTION

Dye-sensitized solar cells (DSSCs) were first introduced by O’Regan and Grätzel in 1991 [1]. DSSCs were more favourable compared to other types of solar cells due to their low cost production, fabrication ease, environmental friendly and flexibility.

Besides that, the reasonably high performance of DSSCs makes them a desirable type of solar cell by the consumers. To date, the TiO2-based DSSCs have achieved the highest power conversion efficiency (PCE) of ~15 % as reported by Burschka et al. [2].

However, alternative metal oxides such as ZnO [3], In₂O₃ [4] and SnO₂ [5] have higher electron mobility than TiO2 that speeds up the electron transport and improves the PCE of the cell.

In₂O₃ is a wide band gap material with a direct band gap and indirect band gap of 3.6 eV and 2.6 eV, respectively [6]. In₂O₃ is rarely used as photoanode material in DSSCs as it exhibits low PCE compared to TiO₂ materials even if it has high electron mobility as mentioned above. Recently, carbon nanotubes (CNTs) have been used to combine with the photoanode materials such as TiO2-CNT, ZnO-CNT and SnO2-CNT to increase the PCE of DSSCs [5, 7, 8]. Introduction of CNTs in the metal oxides improves the electrical conductivity by providing fast electron transfer kinetics which was observed through electrochemical impedance spectroscopy (EIS) unit [9]. Moreover, in terms of morphology incorporation of CNTs induce pores that able to improve light harvesting efficiency in the electrode making an easy pathway for dye-adsorption.

In this paper, we report the DSSC performance of In2O3 and incorporation CNTs with In₂O₃ as photoanode material. There are two types of CNTs which are multi-walled carbon nanotubes (MWCNTs) and single-walled carbon nanotubes (SWCNTs). These two CNTs are used to incorporate with In2O3 nanoparticles. The main aim of this report is to improve the low performance of In₂O₃-based DSSCs by introducing MWCNTs and SWCNTs. The structure, morphology and electrical properties of In₂O₃-CNTs were studied carefully.

EXPERIMENTAL DETAILS

Preparation of Thin Films.

Sol-gel method was used to prepare In₂O₃ thin films via spin-coating technique as shown in Figure 1. All the chemical reagents were purchased from Sigma-Aldrich (USA). 0.1 M of Indium chloride (InCl₃) was dissolved in 50 ml of 2-methoxyethanol to form a stable transparent aqueous solution. The mixture was stirred on a hot plate for 24 h at 60 °C. After the cooling process, the mixture was spin-coated with five layers on a Fluorine-doped tin oxide (FTO) coated glass substrate by using a spin coater (Model WS-400BX-6NPP/LITE). The spin-coated In₂O₃ films were annealed at 450 °C for 30 min in an electric furnace. The same procedure took place to prepare In₂O₃-CNTs except in the first step 0.1 M of InCl3 powder and 0.1 % of CNTs were mixed together and dissolved in 50 ml of 2-methoxyethanol.

Figure 1: Flow chart of In₂O₃ and In₂O₃-CNTs thin films preparation process Fabrication of DSSCs.

The annealed thin films were immersed in ethanolic N719 dye (0.5 mM) for 24 h in a glass petri dish. At the same time, the counter electrode (CE) was prepared by pasting platinum paste on a clean FTO glass substrate via screen printing technique. The CE is then annealed in air at 400 °C for 1 h. The DSSCs were fabricated by sandwiching the immersed photonode thin film and CE together. The DSSCs were assembled by using a parafilm layer and two binder clips with active area of 1 cm². Finally, the electrolyte (Idolyte MPN 100 Solaronix SA) was injected into the cell.

Characterization of DSSCs.

The structure and morphology of the photoanode thin films were characterized by X-ray diffractometer (XRD), atomic force microscopy (AFM) and field-emission scanning electron microscope (FESEM). The photovoltaic performance of the photoanode thin films was analyzed through photocurrent density-voltage (J-V) curve measurement. The electron transport properties were determined by EIS unit (GAMRY Series G300 Potentiostat).

RESULTS AND DISCUSSIONS

The structure of the thin films was confirmed through the XRD patterns. Figure 2 (a), (b) and (c) shows the XRD patterns of In2O3, In2O3-SWCNTs and In2O3-MWCNTs, respectively. The phase structure of In₂O₃ was confirmed as body-centered cubic with lattice constant, a of 10.117 Å (JCPDS no. 01-071-2194. From Figure 2, crystallinity of

the thin films were improved when CNTs were incorporated with In₂O₃ (Figure 2 (b) and (c)). However, when comparing the both CNTs, SWCNTs incorporated with In2O3

showed higher crystallinity than MWCNTs. The XRD patterns of In₂O₃ were attributed at 2θ = 26°, 30.575° and 35.475°, 37.70° and 51.043°, corresponding to the (h k l) miller indices of (211), (222), (400) and (440) planes. Orientation at (222) plane was considered as the preferred orientation [4].

0 50 100 150 200 250 300 350 400 450 500

20 25 30 35 40 45 50 55 60

In te n sity (Cou n ts)

2 Theta (Degree)

(211)

(002)

(222)

(400)

(440)

(100) (004)

--- In₂O₃ Carbon

(211)

(002)

(222)

(400)

(440)

(100) (004)

--- In₂O₃ Carbon

(211)

(002)

(222)

(400)

(440)

(100) (004)

--- In₂O₃ Carbon

(a) (b) (c)

Figure 2: XRD spectra of (a) In₂O₃, (b) In₂O₃-SWCNTs and (c) In₂O₃-MWCNTs Meanwhile, the carbon (C70) patterns were reflected at 2θ = 26.2°, 43.4° and 54.6°, reflecting to (002), (100) and (004) planes. The peak intensities of carbon were very small due to the low concentration of only 0.1 % of CNTs were added in the composite.

The result confirmed that the In₂O₃-CNTs possess a carbonic character in their pore walls. Furthermore, the crystallite size (D) was calculated by using Debye-Scherrer’s formula [11]:





 cos D k

(1)

where, k is the Scherrer’s constant with 0.94 and β is the full width at half maximum

(FWHM) of the Bragg peak. Table 1 shows the crystallite sizes of the In₂O₃ nanoparticles. The crystallite sizes of the In₂O₃ nanoparticles decreased when CNTs were incorporated in the nanocomposite. In addition, Abdullah et al. mentioned that incorporation of CNTs decreases the D [10].

Table 1: Crystallite sizes and surface roughness of the thin films Samples hkl 2θ (°) D (nm) R_a (nm)

In2O3 (222) 30.59 19.21 0.35 In₂O₃-MWCNTs (222) 30.54 16.52 3.54 In₂O₃-SWCNTs (222) 30.51 15.32 6.62

Figure 3: AFM images of (a) In₂O₃, (b) In₂O₃-SWCNTs and (c) In₂O₃-MWCNTs

The AFM and FESEM analysis were done to analyze the morphological characterization of the photoanodes. Figure 2 (a), (b) and (c) shows the AFM images of In₂O₃, In₂O₃-MWCNTs and In₂O₃-SWCNTs, respectively. The surface roughness of the thin films is listed in Table 1. Figure 2 (b) and (c) shows more columnar grain growth which signifies the thin films consist of rougher surface structure. The AFM result denotes that incorporation of CNTs provide rougher surface which improves the short circuit current density (J_sc) of the cell by increasing the light adsorption [12].

Additionally, Uk Lee et al. stated that a photoanode with rough surface structure bounces the incident light that hits the surface and reflects the light indirectly back on the surface of the thin film [13]. Therefore, more incident light is able to be captured by the dye molecules that are on the surface of the films. In2O3-SWCNTs have rougher surface roughness than In₂O₃-MWCNTs as seen from Table 1.

The photovoltaic performance of the DSSCs was measured with light intensity of 100 mW/cm2 under Am 1.5. Figure 4 (a), (b) and (c) shows the J-V characteristics of In2O3, In₂O₃-SWCNTs and In₂O₃-MWCNTs, respectively. The conversion of solar energy to electricity which is the PCE of the DSSCs is expressed in percentage. The photovoltaic performances of In2O3, In2O3-SWCNTs and In2O3-MWCNTs are listed in Table 2. The photovoltaic parameters include J_sc, open-circuit voltage, V_oc, fill factor, FF and PCE.

From Table 2, In2O3-based DSSCs generated low PCE of 0.55 % with low Jsc, Voc and FF of 3.6 mA/cm², 0.34 V and 0.4, respectively. The obtained J_sc and V_oc increased as CNTs were added to In₂O₃ solution. We monitored that, In₂O₃-SWCNTs exhibited higher Jsc of 5.6 mA/cm² compared to In₂O₃-MWCNTs. In₂O₃-SWCNTs attained higher Jsc due to the higher roughness of the thin films by allowing more dye molecules to be adsorbed on the surface of the films. In addition, previous literatures stated that high roughness factor on the surface of the thin films increased dye adsorption in the photoanode layer [9, 14, 15]. However, the Voc and FF of In2O3-MWCNTs and In2O3- SWCNTs did not vary much. This is because, V_oc depends on the nanocomposite conduction band egde and electrolyte which is more or less similar for both samples [16]. Thus, a good interaction between In₂O₃ and SWCNTs happened in the photoanode layer. In conjunction to that, Lee et al. reported that good interaction of metal oxide with CNTs provide longer electron lifetime and inhibits electron recombination in DSSCs [17]. Therefore, the EIS unit was used to determine the electron transport parameters that influenced the overall PCE and Jsc.

Figure 4: J-V characteristics of (a) In2O3, (b) In2O3-SWCNTs and (c) In2O3-MWCNTs

Table 2: Photovoltaic performances of ZnO-MWCNTs based DSSC

Samples Jsc (mA/cm²)

Voc (V)

FF PCE (%) In₂O₃ 3.6 0.38 0.4 0.55 In₂O₃-MWCNTs 4.32 0.45 0.43 0.84 In₂O₃-SWCNTs 5.60 0.46 0.43 1.11

EIS analysis was done in order to analyse the reasons of low performance in In2O3- based DSSCs and higher performance in In₂O₃-SWCNTs-based DSSCs. EIS analysis was conducted under the illumination of 100 mW/cm². Figure 5 shows the Nyquist spectra of (a) In2O3, (b) In2O3-SWCNTs and (c) In2O3-MWCNTs. The Nyquist plots were then fitted by an equivalent circuit model proposed by Mahalingam et al. [4]. The circuit is based on the physical structure of the DSSCs.

Figure 5: EIS plots of (a) In₂O₃, (b) In₂O₃-MWCNTs and (c) In₂O₃-SWCNTs Different frequencies in each impedance spectra represent the photoanode, electrolyte and counter electrode part. The high, medium and low frequency regions represent the counter electrode, photoanode and electrolyte, respectively. The electron transport parameters with DSSCs performance is listed in Table 3 where, R_s is the sheet resistance, R_t is the electron transport resistance and R_ct is the charge-transfer resistance [18,19,20-24].

Table 3: Electron transport parameters of In₂O₃, In₂O₃-SWCNTs and In₂O₃-MWCNTs based DSSCs

DSSC f_max (Hz)

ωmax

(Hz) R_s

(Ω) R_ct

(Ω) R_t

(Ω) τeff (ms)

k_eff (s^-1)

PCE (%) In₂O₃ 476 2989 29.09 35.51 1.54 x 10^-3 0.335 2985 0.55 In₂O₃-MWCNTs 378 2374 28.97 155.7 2.615 0.421 2375 0.84 In2O3-SWCNTs 37.5 236 27.14 676.1 2.682 4.24 236 1.11 Furthermore, the maximum frequency, fmax of the Nyquist plot is the peak frequency at the photoanode region. We can then measure the ωmax since, ωmax=2π f_max. The electron lifetime, τeff, effective electron diffusion coefficient, Deff and electron recombination



^{2 1f}_max



eff 

  (2)

eff

keff



 1 (3)

From the fitted curves of impedance spectra (Figure 5), In₂O₃-SWCNTs achieved the highest τeff of 4.24 ms. In contrast, In2O3-based DSSCs achieved the lowest τeff of 0.335 ms. Consequently, short electron lifetime decreases the J_sc of In₂O₃-based DSSCs to 3.6 mA/cm². In addition, doping of SWCNTs decreased the Rs to 27.14 Ω. Hara et al.

mentioned that, decrease of Rs can improve Voc and FF [25]. In conjunction to that, addition of CNTs in In₂O₃ solution improved V_oc and FF in the cell where value of R_s decreased when CNTs were added to In₂O₃. Besides that, the spectrum of In₂O₃ showed no arc at the high frequency part due to Warburg diffusion of the redox couple of electrolyte. Thus, the low Voc and FF obtained in In2O3 based DSSCs were also due to the missing arc. Moreover, In₂O₃ based DSSCs showed the lowest value of R_ct with 35.51 Ω. Smaller Rct speeds up the electron recombination rate, keff to 2985 s^-1. This event indicated that, more injected electrons were recombined with the electron acceptors that slowed down the electron transport process thus, degraded the PCE of In₂O₃ based DSSCs.

CONCLUSION

In summary, In₂O₃ and In₂O₃-CNTs based DSSCs were successfully synthesized and fabricated by spin coating technique. MWCNTs and SWCNTs were added to In2O3

solution in order to improve the low performance of In₂O₃-based DSSCs. The XRD confirmed the In₂O₃ cubic structure and the crystallite size decreased as the CNTs were incorporated with In2O3. The AFM images also proved addition of CNTs increased the surface roughness of the thin films allowing more dye molecules to be adsorbed on the thin films thus increased the J_sc. However, In₂O₃-SWCNTs possessed smaller crystallite size and rougher surface roughness than In₂O₃-MWCNTs. I conjunction to those factors, In₂O₃-SWCNTs exhibited the highest PCE of 1.11 % with J_sc, V_oc and FF of 5.60 mA/cm², 0.46 V and 0.43, respectively. Moreover, In₂O₃-SWCNTs showed the highest electron lifetime and lowest electron recombination rate of 4.24 ms and 236 s^-1, respectively. Longer τeff in In2O3-SWCNTs caused the Jsc to increase up to 5.6 mA/cm². In addition, the highest R_ct in In₂O₃-SWCNTs slowed down the electron recombination rate and speeds up the electron transport in the cell. Therefore, addition of CNTs in In2O3 solution enhanced the PCE of the DSSCs. However, in comparison to the types of CNTs, SWCNTs exhibited higher performance in In₂O₃-based DSSCs.

ACKNOWLEDGEMENTS

This work was funded by Fundamental Research Grant Scheme (FRGS/2/2013/TK06/UKM/02/3), Photonic Technology Laboratory (IMEN) and

Department of Electrical, Electronic & Systems Engineering, Universiti Kebangsaan Malaysia, Bangi, Selangor, Malaysia.

REFERENCES [1]. B. O’regan, M. Grätzel, Nature 353 (1991) 737-740

[2]. J. Burschka, N. Pellet, S.J. Moon, R.H. Baker, P. Gao, M.K. Nazeeruddin, M.

Grätzel, Nature 499 316-319 (2013)

[3]. H. Abdullah, N.A. Atiqah, A. Omar, I. Asshaari, S. Mahalingam, Z. Razali, S.

Shaari, J.S. Mandeep, H. Misran, Journal of Materials Science: Materials in Electronics 26 2263-2270 (2015)

[4]. S. Mahalingam, H.Abdullah, S. Shaari, A. Muchtar, I. Asshari, Structural, morphological and electron transport studies of annealing dependent In₂O₃ dye- sensitized solar cell, The scientific World Journal 2015, Article ID 403848, (2015).

[5]. S. Mahalingam, H. Abdullah, A. Omar, N.A.M. Nawi, S. Shaari, A. Muchtar, I.

Asshari, Advanced Materials Research, 1107 649-654 (2015)

[6]. A. Raza, O.P. Agnihotri, B.K. Gupta, Journal of Physics D: Applied Physics 10 1871 (1977)

[7]. M.Z. Razali, H. Abdullah, S. Shaari, M.R. Taha, A. Omar, M.A. Yarmo, Journal of Nano Research 28 151-162 (2014)

[8]. A. Omar, H. Abdullah, S. Shaari, M.R. Taha, Journal of Materials Research 28 1753-1760 (2013)

[9]. A. Omar, H. Abdullah, M.A. Yarmo, S. Shaari, M.R. Taha, J. Phys. D Appl.

Phys. 46 165503 (2013)

[10]. H. Abdullah, M.Z. Razali, S. Shaari, M.R. Taha, Electron. Mater. Lett. 10 611- 619 (2014)

[11]. S. Bucak, D. Rende, Colloid and surface chemistry: A laboratory guide for exploration of the nano World, CRCPress, New York, NY, USA, 2014.

[12]. N.P. Ariyanto, H. Abdullah, N.S.A. Ghani, Materials Research Innovations, 13 157-160 (2009)

[13]. S. Uk Lee,W. Seok Choi, B. Hong, Solar Energy Materials and Solar Cells, 94 680-685 (2010)

[14]. Y. Zhang, M.K. Ram, E.K. Stefanakos, D.Y. Goswami, J. Nanomat. 2012 1-22 (2012)

[15]. N. Karst, G. Rey, B. Doisneau, H. Roussel, R. Deshayes,V. Consonni, C.

Ternon, D. Bellet, Mat. Sc.i & Eng. B 176 653–659 (2011)

[16]. G.S.Y. Schlichthörl, S.Y. Huang, J. Sprague, A.J. Frank, Journal of Physical Chemistry B 101 8141-8155 (1997)

[17]. K.M. Lee, C.H. Hu, H.W. Chen, K.C. Ho, Solar Energy Materials and Solar Cells 92 1628-1633 (2008)

[18]. M. Adachi, M. Sakamoto, J. Jiu, Y. Ogata, S. Isoda, J. Phys. Chem. B 110

[20]. J. Bisquert, J. Phys. Chem. B 2 325-333 (2002)

[21]. F. Fabregat-Santiago, J. Bisquert, G. Garcia-Belmonte, G. Boschloo, A.

Hagfeldt, Sol. Energy Mat. & Sol. Cells 87 117-131 (2005)

[22]. Q. Wang, S. Ito, M. Grätzel, F. Fabregat-Santiago, I. Mora-Seró, J. Bisquert, T.

Bessho, H. Imai, J. Phys. Chem. B 110 25210-25221 (2006)

[23]. F. Fabregat-Santiago, J. Bisquert, E. Palomares, L. Otero, D. Kuang, S.M.

Zakeeruddin, M. Grätzel, J. Phys. Chem. C 111 6550-6560 (2007)

[24]. A.B.F. Martinson, M.S. Góes, F. Fabregat-Santiago, J. Bisquert, M.J. Pellin, J.T.

Hupp, J. Phys. Chem. 113 4015-4021 (2009)

[25]. K. Hara, T. Horoguchi, T. Kinoshita, K. Sayama, H. Sugihara, H. Arakawa, Solar Energy Materials and solar cells 64 115-134 (2000)